SENSITIVE GAS CHROMAW3GRAp2IC ASSAY FOE ‘THEi QUA~A~ON OF BRETYLHUM IN PLASMA, URINE AND MYtxxRmAL TISSUE

EUGEXE PATTEBSON, PHILIP STETSON and BENEDICT E. LU-

Departmint of Pharmacology C& The Upjoim Center for CIirtical Pharmacology, The Unitmsity ofc%i&~m Medical School. Ann Arbor. Nick 48109 (U.S.A.)

(First received April 17th. 1979; rev&d manusa5pt received August 6th, 1979)

SUMMARY

A sensitive analytical method has been developed fcr the quantitatioa of bretylium in plasma, urine and myaardial tissue. Bretylium and the internal standard, UM-360 (o- iodobenzykimethylaonium), are extracted and isolated as the iodide salts_ Sodium benzenethioMe is added and the mixture heated to 100” for one hour. This results in the formation of 2-bromobeozyl phenyl thioether and 2-iodohenzyl phenyl &k&her, which GUI be rieparated and quantitated by gas chrom&ographg. Good reliabii~ and repradueibil- ity can be obtained using ele&ron+apQre detection with quantities of bzetylitxm as smalX sslng.

SNTRODUCTXON

BretyIium tosylate (bretyio& o-bromobenzyletiay~dimethylamnonium tosyl- ate) is a unique antiazrh~c agent p ossesskg antif~brillatory actions [1,2] . The phammcokimtic properties of bretylium are poorly described due to the lack of an emlytical method with both specificity and sensitivity for the qt_&emay. An ear&x method used to messrrre bretylium in urine [S] did not possess the sensitiviQ~ needed to measure bretylium levels in plasma. The

’ previous method was based upon the ability of the quaterzm. to bind methyl ormge, and lacked spe&icitg for bretylium. The method used in the present study for qua&St&ion of bref@ium is based upon a procedure used for quan- tit&ion of acetylcholine 141 as modified for use iu quaMtatig bretylium by Km- et al. [5]. The procedure involves the removal of the o-bromo- bemylgroup from bretylitm by sodium bexizenethiolate and the formation of an o-brornobemyl phenyl thioetber which can be measured with excellent

*To ~&om correspondence should be addressed. : .

34

smsititi~~ by electron-capture g=phy -

detection a!Cter separation by gas chroma’i~

Brei;ylium and UM-360 (o-iodobmzyltrime~%ylammonium) se extzacted and Isolated from samples as the iodide tit. ,Qddition of scdium benzene- thiolate results in the formation of o-bromobenzyl phenyl thio&her and o- iodobenzyl phenyl thioether derivatives which can be separated and quantita& ed by gas chromatography.

--

A gas chromatograph (Hewlett-Packard Model 76lOA) equipped with a 63Ni ekctron-capture de+etor was used. Glass columns (1.83 m X 4 mm I.D.) with the followkg packings were employed: 4.3% OV-161 on CasChrom Q (100-120 mesh), 5.9% QV-1 on Gas-Chrom Q (80-100 mesh), 3% OV-17 on Gas-Chrom Q (100-120 mesh), and 4% GV-225 on Gas-Chrom Q (109-120 mesh) (Applied Sciesce Labs., State College: Pa., U.S.A.). Samples (0.1-2 ~1) were :kjected manually using standard commercial micro,rgringes (Hamilton, Reno, Nev., U.S.A.). The following temperatures were used: injection port, 250”; column oven, 190”; and electron-capture detector, 250”. The carrier gas, argon-methane (X9:1), flow-rate was 56 ml/&.

A Finn&an gas chromatograph-mass spectromc’er (electron impact .mode) was used for identification, of the major chromatographic peaks. Helium was usedasthecarriergas.A3% OV-17 on Gas-Chrom Q (100-120 mesh) (1.7 m X 2 mm I.D.) cohunu was used. Injection port, column, and detector temper- atures were as described above.

Extraction

Brdylium and UM-360 were extracted from biological fIuids as the iodide salt; using the method of Vidic et al. [S] as modified by Pohlmann aud Cohen 1’73 _ EM-360 chloride, 50 ~1 of a 5.5 pg/ml solution, and 1 ml distilled water

-were added to: I ml plasma cr serum, 1 ml dissolved tissue sample, or 100~1 urine. Chloroform (3 ml) was added and the tubes were vortexed for IO set followed by centrifugation at 1500 g for 5 min_ The chloroform layer wes removed and discarded. Potassium triiodide [iodiue-potassium iodide-water (1:2: 10 w/w/v) J , 200 yl, and chloroform, 3 ml, were added. Tke svnples were shake9 gently on a reciprocstig flak bed shaker (100 rpm) -Zor one hour fo!hzw- ed by centrifugation at 1500 g for 5 min. The chloroform layer was transferred to a conical glass vial (Reactivial, Pierce, Rockford, Ill., U.S.A.) and evaporated to a dry residue under a stream of. c&y nitrogen, The presence of elemental iodine in *the &Y residue does not interfere with subsequent derivatiiation and quantitation although it can be removed by the addition of 50 ~1 of 16% ascorbic acid in methanol followed by vortexiug and evaporation of the meth- ZUlOi.

Dekmfizatiun

83diilm benzenethiolate (3 mg/ml in ethyl acetate), 200 ~1, was added and

35

the vial capped securely. The samples were Mhzxed in a sand bath heated to 100-120” ‘for one hour and allowed to cool at room temperature. The ethyl acetate was removed uhder a stream of dry nitrogen and the residue dissolved in 100 ~1 hexane or cyclohexaue. Injection of the ethyl acetak ;iftes derivatiza- tion and cooling to room temperature reveals the presence of late peaks which interfere with subsequent injections. These peaks were removed by evaporazion of the ethyl acetate and dissolution of the residue in cyclohexane or hcxane,

Preparation of samples for analysis M&e mongrel dogs weighing 14.0-16.3 kg were anesthetized with intrave-

nous sodium pent&barbital, 30 mg/kg. A left external jugular vein cannula was inserted for drug administration and a second cannula was inserted into the in- ferior vena cava through the right femoral vain for withdrawal of blood samples for bretylium analysis_ Blood was wi_thdrawn into commerciaUy prepared hepsrinized containers (Vacutainer, Be&on-Dickinson, Toronto, Canada). A bolus of bretylium tosyla&e (6 mg/kg) was administered and plasma samples ob- tained at appropriate intervals. The animals (;L = S) were sazificed 12 h after the administration of bretylium and 300-350-mg sections of atrial, left v~- tricular, and right ventricular myocardium were removed for analysis of bretyl- ium concentration_ A 12-h urine sample was removed from the bladder. The myocardial samples were dissolved in 500 ~1 of 12 N sodium hydroxide with 800 ~1 OL 7 N hydrochloric acid added after tissue dissolution. The above ex- periments were designed to demonstrate the ability of the assay to detect bretylium in biological fhzids and tissue.

RESULIYS

C&antifati& of bretyiium and e;lttractiim efmiemy Electron-capture detection was used for the measurement- of the haloge-

nated thioethers. Typical chromatograms obtained from injected standards and plasma are shown in Fig. 1. Peak height ratios (o-bromobenzyl phenyl thio- ether:o-iodobenzyl phenyl thioether) were used for quantitation. Standard curves were prepared for plasma, urine and tissue by the addition of lnown qmtities cf br&ylium to the samples. A linear: correlation exists for samples representig bretylium plasma concentrations of 25-3000 ng/ml; r = 0.998. Linearity also exists for standards representing as little as 1 ng/ml bretylium or as much as 20 pg/ml bretylium; I-20 ng/ml (r = 9.992) and 2-20 &g/ml (r = 0.995). The quantity of UM-360 added to the last two standards was modified to facilitate measurement of peak heights_ The injected sample volume was varied so the detector response ranged from O.I-2.0. PO-‘o A and retained a linear response ratio. Standard curves for urine, IO-200 pg/ml (r = 0.999) and myocardial tissue, 1.20 pg/ml (I = 0.990) also were linear.

Bretylium, 25-SO00 ng, and UM-360, 250 ng, dissolved in methsno’l, were added to conical glass vials aud the methanol evaporated under a stream of dry nitrogen. These samples were run in parallel with p&ma bretylium s+&ndards representing 25-8000 ng/nrl bretylium. IJM-360, 250 ng, dissolved in metha- nol, was added to the dry chloroform extract of plasma and the methanol evaporated under dry nitrogen. Recovery of bretylium *horn plssmz uz%ng

36

Pig. I. GSA ctiromatograms from ~&mdards and plasma Sample A vas prepared horn stock sohtiocs of bretylism and TJM-360. Peaks: 1 (setentiorn time 2.9 mire) = benzenethial di- sulfide, a coa*k.minsut formed f&n sodium henzenethiofate derivatization; 2 (retention %ime 3.6 min) = o-bromobenzyl phenyl thioether, a derivative of bretylium; 3 (retention time 5.3 min) = o-iodobenzyl phenyl thioether, a deGvative formed Crorn UT@360.

potassium tziiodide extraction as described previously was 95.9 + 3.0%

Mass s~eci-roscop~ and iderttification of major peaks A total ion chromatogram and mass spectra of the three major peaks are

given in Fig. 2. Peak 1 of tbe toti ion chromatogram, with a base peak (m!e = 109) cOrresp~n&ng to clezwage of the d&Side bond and a second peak (rr,/e = 218) it?~~Xltig the molecrrlar SpecieS Was shown to be benzenethid did-

fide. Feak 2, with a bzse peak (m/e = 171) corresponding t% fragmentation of the thioetier with formation of the resonance stabilized bromotropylium ion and a second peal: representing the molecular species (m/e = 280) was shown to be o-bromobznzyl pbenyl thioetier. Peak 3, with a base peak (m/e 217) corre- sponding to formaGon of an iodotropyliun ion by fragmentation of the So- et&r, was shown to be o-iodobezzyl phenyl thioether.

Carhe plcrsm, urine and myocardiai t&sue samples Plasma levels of bretylimn taken at appropriate intervals over a period of

12 h after 2 bolus intravenous injection .of bretylium tosylate, 6 mg/kg, are given in Fig. 3. ?lasma bretylium leveis varied from a maximum of nearly 20 yg/ml Fast after bretylium administration to a minimu& of just over 2 &ml at 12 h,

Myocardial tissue levels of bretylimn were measured at the time of sacrif&, . 12 h after the * - a ’ tion of breelimn (Fig. 4). Atria;l t&sue levels of bretyl-

i-mn were lower than those seen in right gnd left wenfzicular myocttrdim. A seven-fold ratio of ventricular myocardM tissue to plasma, bretylium con- centratioti &g/g myocardiai tissue:clg/ml plasma) was seen.

m/e =217 MW=32c3

. I- L I lm IEn an 2se Fig. 2.M.zs5 spectrohcopicaaalysis.TEettwemajorpeaHsmezsur.~ by electronimpactmzs spectroscopY corresponded to the three peaks measured by electrori capture in Fig_ 1. 1. (m/e=l09. m/e=228), benzenethiol &sulfide; 2, (m/e=l?l, m/c=109, na/e=280), o-bromo- be~lp~enylthioether;3,(m/e=217,m/~=109)o-iodobenzylpheny~thioether.

37

, MW=280

A 12-h urine sample was coI.lec.~ and theexcreted bretyliun quantitated. The total 12-h excretion of bretyfiwn was 23.4 f 3.9 mg (50.6 + 8.3% of the administered dose).

All samples were performed in duplicate with a mean difference of 3.8 + 1.8% between samples.

2Yne method af Kuntzn~an et. al. 151 has a nrunber of problems which prevent its general use for pharnwokinetic studies. (1) A large numbeE of sample transfers and extzactions are necesazy for the isolation of bretyliwn "aad'the ~eliWn&ion of substances from plasma &xd urine with shniktr retention

38

i 0 2 4 6 6 10 12 so LEFT ATRIA RIGHT

HOtlRS 9zs-f Lv. eOLUS VENTRICLE VENTRICLE

Fig_ 3. Plztsxm b~t@illm concentrations. Piasma bretylium concentrations in an k&ketiz- ed ckg after ezx ultzave;lous (i-v.) bolus injection (6 mg/kg) of bretjrfium to&ate (mean f s’%ndacd error of the mean).

Fig. 4. Mywardial tissue concentrations of bretylium. Eretylium concentrations k~ catie myocardizd !&aue. 12 lo after a &gle intravenolls bolus injection of 6 mg/kg bretyhm ckrsylate (mean + standard error).

t%nes to +&ose of the derivatized product and external s+andard. (2) Qua.n%a- tion of the o-bromobenzyl phenyl tbioether is possible only by the addition of an external standard, occurring after the extmction and isolation of bretylium. (3) The sznallest detectable quantim of bretglitzn, 70 ng, is greater than the plasma levels necessary for observation of. linear first-order pharmacokinetics in man. (4) The poor chromatographic separation of o-bromobenzyl phenyl thioether from -benzenethiol required the use of conditions which resulted in long reten$ion times for the o-bromobenzyl pheriyl thioether (14 min) and the extemal standard (18 min)_

The assay procedure for bretylium as given in this paper overcomes ;nany of the problems occurring with the previous method. A single sample wash, followed by chloroform. extraction of brewlium as the triiodide salt, is the only procedure necessary for the isolation of bret@ium in a relatively pure form. Derivatization foliowed by evaporation and dissolution of the haloge- natzd thioethers yields a clean chroma~ with no interfering peaks. This is in contrast to the multi-step procedures previously used to isolate bretylium in a reL&vely pure form [3] and to obtain a derivatked product with sufficient purity to achieve a clean chromatogram 151.

The use cf an internal staFdard undergoing identical isolation and derivatiza tion procedures as the measured product,- bretylium, eliminafes the problems which can occur due to inaccurate volume transfers occurring prior to the addition of an external standard. Also, the addition of an internal standard serves as a control for both the exkaction procedure and derivatization proce- dure. The failure to detect breI@ium in a sample, therefore, can be attribukd to the absence .of bretylium in a specimen, thus eliminating the possibtities of hu.By ex&a&ion or derivatizzttion.

The assay is both specific and sensitive for breQ$&m. Quantities of breel- ium in excess of 1 ng are measured easily. The bretylium &sma levels in the

39

anesthetiied dog were considembly higher than are seen in humans at similar dosages of brekykm [5]. However,- we stil.I have shown the abiliw to measure bret@un in quantities as small as one nanogram from plasma s&daros.

Benzenethiol and o-bromobeuzyi phenyl thioether were easily separated by a number of cohnnn packings (see Apparatus) with retentioti times for the u- bromobenzyl phenyl thioether of 3.2-5.0 min and for o-iodobenzyl phenyl thioether of 4.2-6.2 min. No peabs with higher retention &es were seen_ This allows relatively rapid quantit&ion of injected samples.

The above assay possesses the required specificity, sensitivity and ease of sakple preparation needed for pharmacokinetic studies of bre@lium in man and animals.

ACKNOWLEDGEMENT

This study was supported by a grant from the United States Public He&h Service HLBI-HL-05806-19.

REFERENCES

J.A. Cooper axed J. Trieden, Amer. Heart J., 82 (1971) 703. J. Koch-Weser, N. EngL J. Med., 300 (1979) 473. W-G. Duncombe and A.. McCoubrey, Brit. J. Pharmacol., 15 (1960) 260. D.J. Jenden, I. Hanin and S.L. Lamb, Anal. Chem., 40 (1968) 125. R. Kuntzman, L Tsai, R. Chang and A.& Conneg, Clm. Pharmacol. Ther., ll(l970) 829. H.J. Vidic, H. Dross and H. Keivitz, J. Klin_ Chem. Klin. Biochem.. 10 (1972) 156. J.L.W. Pohlmann and S.L. Cohan, J. Chromatogr., 131(1977) 297.